WHAT
front page
entry archive
my flickr photos
professional home
my twitter
my wishlist
my tumblog
my facebook
WHO
Another MRTG Chart
2010-07-06 : rambling
The weekly internal temperature chart is looking even worse today, so when my box fell off the grid, I assumed there was a thermal cause:
Turns out there wasn’t a direct thermal issue (though the battery controller on the RAID controller is pretty ornery about how hot the battery got). How do we explain the story?
Here’s line voltage into the UPS that feeds my equipment rack:
And UPS charge levels, for added effect:
And that’s pretty much the story. I’m guessing there were some crazy power issues today trying to cool Connecticut when it was running over 100 degrees.
Everything went into safe shutdown once the UPS ran out of power, and then turned back on once the power was back. Sort of. My core switch didn’t come back online; I think there’s a timing issue about when it first powers up, and one of the switches on the network never ran out of power (it has a few hours of battery life on its UPS).
Anyhow, having a girlfriend willing to go check all of this stuff out when you’re on the other side of the country is pretty awesome.
Two questions:
1) What the hell scale is your temperature graph in? Is the peak really 4800 (4.8k) * 100 celsius? You melt steel with that thing?
2) Why was your UPS hemorrhaging charge before you line voltage dropped off?
— Bill 7 July 2010 #
Sorry about that; the scale is off, 4.8 = 48 degrees C. I always knew what it meant so I never fixed it.
As to hemmorrhaging charge, it was probably seeing line voltage variations and buffering the output voltage; it does a pretty god of smoothing things out, unlike the other more primitive UPS boxes in the apartment.
— Aaron N. Tubbs 7 July 2010 #